Our group of investigators has identified and confirmed a novel chromosomal region, 9q22.2-31.2, that demonstrates evidence for linkage (P=0.00016) to familial colon neoplasia in a pattern consistent with dominant autosomal disease alleles in this region accounting for an estimated 40% of colon cancers and advanced colon adenomas in the colon neoplasia families we studied. Our initial finding was the result of linkage analysis of a sibling pair study that was designed to identify novel human colon cancer susceptibility loci based on an unbiased whole genome scan employing sibling pairs drawn from 53 "study-4" kindreds in which 2 or more siblings had by age 65 or younger been affected by colon cancer, or affected by advanced colon adenomas (of size >1cm or with histology demonstrating high grade dysplasia) that are direct cancer precursors. Our initial discovery has now been confirmed by us in an enlarged cohort of 120 "study-4" kindreds, as well by independent investigators in Britain and in Scandinavia. After fine mapping, our currently refined linkage region spans 8.8Mb. The main purpose of this new proposal is to further narrow this candidate region using non-linkage based techniques, such as i) SNP association, and ii) mapping gene regions that familial tumors target for loss {or for gains};followed by then identifying the actual novel disease gene and pathogenetic disease alleles within the 9q22.2-31.2 interval. The specific steps we thus now propose are: i) To analyze the 9q22.2-31.2 interval for evidence of either a tumor suppressor gene {or an oncogene} by examining this interval for, respectively, either loss of heterozygosity (LOH) {or for increased gene copy number} in the tumors arising in kindreds that show linkage to the region. To then further narrow this genomic interval by defining the minimal region of LOH {or of increased gene copy number} that is shared in common among these familial colon tumors. ii) To examine disease association among {a comprehensive panel of 5147 SNP polymorphisms, with average spacing of average 2.7kb, that capture both the common and the rare genetic variation across the 9q22.2-31.2 linkage interval;} thereby directly or indirectly mapping the location of any potential founder disease allele that would account for tumor development in many or all of these colon neoplasia families. iii) To directly look for potential disease genes and pathogenetic disease alleles by studying our best linked kindreds via high throughput sequencing of all genes within our maximally narrowed 9q22.2-31.2 interval, examining a maximum of 102 genes contained within our currently narrowed 8.8Mb linkage interval. iv) To in our best linked families construct "converted clone" human-mouse somatic cell hybrids to capture individual parental human chromsome 9 copies. To employ these captured haploid parental chromsomes to test the genes in our maximally narrowed interval for disease causing mutations that would not be detected by standard approaches, including testing for exon deletions and testing for mutations that cause exon skipping. Success of this project would have substantial implications for public health. PUBLIC HEALTH RELEVANCE: Colon cancer is the second leading cause of cancer death in the United States, with 20% of colon cancer patients reporting a positive family history of colon cancer also affecting either a parent or sibling. Identifying a new gene that is a common cause of familial colon cancer would allow identification of individuals at high risk for this disease, for whom aggressive screening and early detection could be life saving.